Some Unattractive Meta-Ethical Positions, Free to a Good Home
06 Sep 2024 15:50
One of my vices is reading meta-ethics, i.e., philosophers arguing not about what is right and wrong or good and bad, but about what kinds of things count as good and bad, and how we could know. Like most of my vices, it is idle, yet not something I will give up any time soon. But it has its frustrations (also like most of my vices...); one of them is that it seems to show a severe lack of logical imagination. So here are some meta-ethical positions which do not seem to me to be obviously self-contradictory, but also notably absent from the philosophical literature. I am bad at naming things, so I'll just number them as they occur to me.
- Objective ethical truths exist, but they involve concepts which are intrinsically beyond human comprehension. By analogy: it's pretty well-established that many animals have some kind of sense of number, and are able to grasp "more" or "less" or "equal". (See, e.g., Dehane's The Number Sense.) But it seems absurd to think that those animals have the concepts to grasp the (pretty objective-seeming) truths of arithmetic, and not just because nobody has taught them. Like animals which have evolved a dim numerical sense, we may have evolved a dim ethical sense, while simply being organically incapable of grasping anything more. Murdering kin who have done us no harm doesn't sit well with most of us, and this may be objectively true, just as it's true that \( 3 < 4 \). Thinking that therefore we must be able to figure out ethics is like expecting that, because a squirrel knows to prefer 4 nuts to 3 nuts, it can grasp Fermat's last theorem.
- Objective ethical truths exist, but they involve concepts which nobody has formulated yet. We are not so much in the position of a squirrel trying to grasp arithmetic, as of our own ancestors with dim notions of number and space, before the long, slow process of cognitive development that led to ideas running from "addition", "angle", "area", etc. to "zero". These are all notions with histories, requiring many long ages of both daily use and occasional genius to be shaped into the tools which do let us grasp mathematical truths. Expecting our current moral and ethical vocabularies to be up to the job is like thinking that we should have been able to do mathematics as soon as we learned to count "one, two, many" on our fingers.
- Objective ethical truths exist, but they involve concepts we have not yet
formulated, and whose connection to the intuitive concepts of our everyday life
is at best remote. Here the analogy is not mathematics but physics: we start
with observing medium-size dry goods (as the saying goes), and try to explain
what happens using concepts like "solid body" and "hit" and "push", but it
turns out that most of what we call solid bodies is empty space, "hitting" and
"pushing" are really explained by electromagnetic interactions (even when no
amber or lodestones are anywhere in evidence), and explaining why that
doesn't lead to matter disintegrating in an ultraviolet
flash requires quantum
mechanics, including the Pauli exclusion principle. It might be that one
of our current ethical theories in the fortunate position of
being merely as wrong as ancient atomism
and/or the 17th century mechanical
philosophy, i.e., false in every detail, but a fruitful error which in some
respects roughly approximates a more adequate set of concepts and theories,
and might get there if both applied ruthlessly and held to high
standards of clarity and explanatory adequacy. Under this interpretation,
then, the best existing ethical theorizing should remind us
of Lucretius coming up with seven different
atomistic explanations for
thunder, starting
with clouds scraping against each other (VI 96--106, A. E. Stallings
translation [London: Penguin, 2007], p. 199):
First, the reason that the thunder shakes the azure sky
It should remind us of this, only with less literary value, and with the possibility that the equivalent of atomism might well not have been invented yet.
Is that the scudding clouds crash into one another high
Up in the aether when the winds are warring, for no sound
Comes out of a clear patch of sky; but wherever clouds are found
Mustered in thicker ranks, it's from that corner that the rumbling
Thunder's usually more apt to mutter mighty grumbling.
Besides, the stuff of clouds cannot be dense as wood or rock,
Nor can they be as wispy as the mist or wafting smoke,
Because they'd either sink like stones, sagging beneath dead weight,
Or else like smoke, failing to cohere, would dissipate,
Unable to contain the snow or keep the hailstones pent. - Objective ethical truths exist, and involve concepts we can grasp, but they are vastly more complicated than can fit in our minds. On Kolmogorov complexity grounds, then, fragments of them which are small enough to fit into human minds will tend to seem bizarre and random, perhaps ritualistic or supersititious. (E.g., if, but only if, one sees exactly three birds flying to the west at dawn on the summer solstice, attacking a family member who has never harmed you is not only permitted but actually obligatory.)
- Objective ethical truths exist, and human beings can comprehend them, but we are so constituted that we can never believe them, or act upon them. I leave open whether we are automatically to be condemned upon these grounds.
- Objective ethical truths exist, but they imply that human life in all its forms is utterly, irredeemably vile. Consequently, not only does acting ethically not making anyone happier, or improve anyone's life in any way, it is right and proper that it does so, because we are inherently contemptible.
- (6), but combined with a clause saying "with the following exceptions", where the exceptions fall under any of the preceding theories. (This is how I'd go if I wanted to make any of these into a cult. Kolmogorovian Gnosticism, in particular, would seem to have potential.)
- There are objectively true statements containing ethical terms, which are true in all (valid) ethical systems, but these truths are not first-order ethical statements about what is right or wrong. The analogy here is to special relativity, where the time \( t \) or distance \( r \) between events is relative to reference frames, even though the "interval" \( c^2t^2 - r^2 \) is invariant. (If the interval is "spacelike", i.e., negative, the order in time of the two events is even reference-frame dependent.) This analogy doesn't push things quite as far as I'd like, because within a given reference-frame the time interval \( t \) and distance \( r \) are perfectly objective facts. But it does, I think, manage to hint how the objective truths might be only about how various ethical claims must relate to each other.
- Act consequentialism is true, but to evaluate an act requires
evaluating all of its causal consequences, no matter how tenuous or
remote, or how uncertain or unforeseeable from the viewpoint of the actor,
going forward until the future light-cone of the act is empty, or at least
empty of morally-consequential beings and events. Ethical evaluation is thus at least
P-complete and probably
much worse. ("Probably"
because, based on the
literature, I am pretty sure we could devise a reduction
to SATISFIABILITY but I haven't actually done it.) There thus is an
objectively correct action to take in all circumstances, but no way of knowing
what it is, for any creature less providential than Laplace's "vast and
considerable intellect". This does not, however, offer any excuse to less
foresightful actors. They might happen to act correctly, but only by
accident. --- Now you might say that if we mere mortals cannot actually foresee the consequences of our actions, we cannot be judged at fault for them or expected to follow consequentialism, that "ought implies can". To which I respond "I do not see the necessity".
(To be truly nerdy, relativity creates an issue even for the vast and considerable intellect, which I will dub "Vaci" for short. Suppose Vaci to b mulling over how to act at a point-instant \( \mathbf{x}_0, t_0 \), using coordinates from Vaci's own reference frame. What's going on at any point-instant in Vaci's future light-cone, say \( mathbf{y}, \mathbf{t} \) with \( t > t_0 \), will depend on the contents of the past light-cone of \( mathbf{y}, \mathbf{t} \), and this includes point-instants outside the past light-cone of \( \mathbf{x}_0, t_0 \). But Vaci can have no causal knowledge of what is happening in those regions of space-time.) - Ethically significant beings have not yet evolved, and will not do so for millions or billions of years --- perhaps Stapledon's Fifth Men, or Eighteenth Men will qualify, but we do not any more than Australopithecus, or Dimetrodon. The world is thus now ethically neutral, except to the extent that what we do might matter to those future beings.
- There was a time before color vision evolved, and there will be a time after the last creature with color vision has died out. (This is true even if we make every allowance for vagueness in "color vision".) Analogously: Consequentialism is true, but only consequences to trilobites mattered. The world was ethically neutral before the Cambrian explosion, and has been ethically neutral since the end of the Permian mass extinction. (Whether perfectly-revived trilobites would somehow be ethically significant again would be another interesting question.)
All of these are rather grim, but I don't see why that should be disqualifying. If you've run across them in the wild, please let me know. If you want to take them up, feel free.
(First version: 14 Mar 2018; revision 17 Apr 2022, small tweaks 26 May 2022, 5 September 2024)